Treatment Techniques for Removing Natural Radionuclides from Drinking Water (TENAWA)

This document is borrow from, the TENAWA report 

Contract N^o: FI4P-CT96-0054

Nuclear fission safety programme of the European Union (1.1.97 to 30.6.99) Introduction

The reason to initiate the TENAWA project was the fact that in several European countries ground water, especially bedrock water, may contain great amounts of natural radionuclides, derived mainly from the ²³⁸U-series. Elevated levels of natural radionuclides in ground waters are mainly associated with uranium and thorium rich soil and rock minerals, or with uranium, thorium and radium deposits. Countrywide surveys of natural radioactivity in drinking water have been conducted in several European countries. The surveys made, e.g., in the Nordic countries especially indicate that high concentrations of radon and other radionuclides usually occur in water from wells drilled in bedrock. In surface waters the concentrations are usually low as in ground waters occurring in soil deposits.

In most European countries ground water is widely used as a raw water source for water works. There is also an increasing tendency to replace surface water with ground water. However, this involves an increased risk of natural radionuclides in water.

Elevated levels of natural radionuclides in drinking water are accompanied with potential health risks for the population by increasing the radiation dose. Therefore, water should be purified before using it. Various processes based on different principles can be applied to remove the radioactivity from water. Aeration is a method that is usually applied to remove radon (²²²Rn) from drinking water. Aeration should be used if the concentration of radon is high, whereas the granular activated carbon (GAC) filtration can be used when the radon concentration of water is moderately low. Ion exchangers are mainly used to remove uranium (²³⁸U, ²³⁴U) and radium (²²⁶Ra). They can also be used to remove lead (²¹⁰Pb) and polonium (²¹⁰Po) but this needs to be studied in more detail. Membrane techniques, such as the Reverse Osmosis (RO) or the nanofiltration (NF) are capable of removing uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po simultaneously. Different natural radionuclides can also be removed using various types of adsorptive filters. Some equipment originally designed for removing iron (Fe) and manganese (Mn) are capable of removing also natural radionuclides.

When different kind of treatment methods are used to remove natural radionuclides from drinking water, wastes containing these radionuclides will be produced. When in use the GAC filters can increase the dose to the residents if not properly shielded or installed.

Objectives

The overall objective of the TENAWA project was to study various removal methods and equipment currently commercially available, and their ability to remove natural radionuclides from drinking water by carrying out laboratory and field experiments. The measurable objectives of the TENAWA project were as follows:

• to make recommendations on the most suitable methods to remove ²²²Rn, ²³⁸U, ²³⁴U, ²²⁶Ra, ²²⁸Ra, ²¹⁰Pb and ²¹⁰Po from drinking water of different qualities (soft, hard, Fe-, Mn- and humus-rich, acidic, etc.)
• to test various commercially available equipment and their ability to remove radionuclides
• to find new materials, absorbers and membranes capable of removing radionuclides effectively
• to issue guidelines for the treatment and disposal of radioactive wastes produced during water treatment.

Results

A literature review with the title: "Natural radionuclides in drinking water in Europe and treatment methods for their removal" was prepared. The main potential risk areas for the occurrence of high contents of natural radionuclides in ground and surface water in Europe are pointed out and data on the natural radioactivity in drinking, mineral, ground and surface water from 17 European countries are presented. An overview on the possible treatment methods to remove natural radionuclides from drinking water is given and human health aspects as well as the regulations regarding natural radionuclides in drinking water are presented. The review offers a good basis for the future studies concerning natural radioactivity in Europe.

Eleven aerators planned for radon removal were studied. Different type of installations were also studied. The most efficient aerators in this study were "Radonett" by Sarholms Ab and "Radox" by Overcraft Oy. Good removal efficiencies were attained with the RF-series aerators by Oy WatMan Ab but the removal efficiency attained with the "Orwa" aerator by Vartiainen Oy was clearly lower.

The studies on radon removal in small water works comprised the assessment of radon removal efficiencies of various aeration techniques in 45 Finnish, Swedish and German water works. The results of those water works where the aeration method was originally designed either for radon or carbon dioxide removal, showed that the radon reduction varied from 67 to 99%. Most of the radon is also removed if the aeration is applied for iron or manganese removal, whereas low radon reduction is attained when conventional water treatment processes, such as lime filtration, water softening or ion exchange, are used.

Radon removal by the granular activated carbon (GAC) filtration was studied in field experiments. The main objective was to investigate radon removal by the GAC filtration in the domestic use. Radon can be effectively adsorbed by the GAC filtration. The short-lived decay products of radon are also retained in the filter and possibly uranium, ²²⁶Ra and ²¹⁰Pb. As a consequence the filter will emit gamma radiation. The external gamma dose rate on the surface of the filter can be up to 100 µSv/h. The radioactivity of the spent carbon can also be a problem when it is disposed of. The results showed that radon was efficiently removed by most filters. Seven units out of thirteen were capable of removing more than 99.9% of radon. The lowest removal efficiency observed was 92.9% and it was possibly due to the elevated uranium content of the water. Besides radon, the GAC filters were capable of retaining various amounts of uranium, ²²⁶Ra, ²¹⁰Pb, ²¹⁰Po and radon progeny. Therefore, the spent GAC batches may contain several hundreds of kilobecquerels of ²¹⁰Pb.

Batch experiments carried out in the laboratory with seven different types of GAC filters showed a high but not uniform efficiency also for removing uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po. The results showed that adsorption of uranium and ²²⁶Ra depends strongly on the carbon type, because smaller granular size results in a larger surface and higher contact time. The adsorption of uranium and ²²⁶Ra depended also on water hardness, dissolved organic carbon (DOC) and pH, while ²¹⁰Pb and ²¹⁰Po were removed quantitatively and independently of these factors. Also column experiments with a commercial filter system were carried out to verify the results of the batch experiments. The results for uranium and ²²⁶Ra agreed quite well with the results of the batch experiments but not for ²¹⁰Pb and ²¹⁰Po, which were not removed quantitatively nor independently of various factors as in the batch experiments. The removal of uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po was also studied in field tests by using the same GAC filters that were used in the radon removal studies. The results of the field tests indicated that the GAC filters were not capable of removing all these radionuclides simultaneously. The removal efficiencies of various nuclides varied in a quite large range in different test locations. The GAC type used in field experiments is not appropriate for removal of long-lived radionuclides.

Commercially available iron (Fe) and manganese (Mn) removal equipment were studied in field experiments. The aim of this study was to find out if the equipment originally designed for Fe and Mn removal, could remove also natural radionuclides. The commercial iron and manganese removal equipment are based on three main principles: aeration-filtration, greensand filters regenerated with KMnO₄ and ion exchange. The field tests were performed in 20 private households. Most of the iron and manganese removal equipment available on the Nordic market were tested. The removal efficiencies for the different radionuclides varied within a large range. For radon, the equipment based on aeration-filtration were the most efficient (reductions from 12 to 89%). Uranium and ²²⁶Ra were best removed by ion exchange techniques (reductions from 50 to 99%) when both anion and cation resins were applied. Removal of ²¹⁰Pb and ²¹⁰Po varied within a large range by various equipment mainly due to their speciation.

A great emphasis was put on ion exchange technique. Besides summarizing the available information about ion exchange for the removal of natural radionuclides from drinking water, several batch and small column experiments and field and laboratory tests on commercially available systems were conducted in order to find the types of resins best suitable for the removal of uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po. The aim was also to study the influence of different raw water qualities on the removal process, to examine the regeneration process and to evaluate the quantity and quality of waste produced by this technology. Strong basic anion resins for the removal of uranium and strong acidic cation resins for radium removal performed best. The efficiency for ²¹⁰Pb and ²¹⁰Po varied a lot, since the main proportion of these nuclides is supposed to be particle-bound in natural waters, and therefore no ion exchange process in the real sense, but adsorption to the resins is responsible for their reduction.

Adsorption technique is used today for the treatment of surface water and ground water. This study focused on finding new absorptive materials for the absorption of the non-volatile elements, uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po. Hydroxylapatite was found to have a good capability to adsorb uranium and ²²⁶Ra, but its application in domestic use would require the development of a stable filtration mass. Four different reverse osmosis (RO)- and one nanofiltration (NF)-system, typical from a great number of commercially available ones, were tested in laboratory experiments. The devices removed in average from 95.6 to 99.8% of uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po. No significant differences in removal efficiency were observed between the RO-units and the NF-system. In the NF experiments at a plate module pilot plant the five most important uranium species for the mobilisation of uranium in natural water were generated in different model waters. Their rejection was determined at six NF membranes and at two open RO membranes. The uranium rejection of the NF membranes varied from 95 to 98% in most cases. The two RO membranes rejected from 98 to 99.5% of uranium.

A literature survey on speciation of natural radionuclides in ground water indicated that very little is known especially of speciation of ²¹⁰Po and ²¹⁰Pb in groundwater. The presence of ²¹⁰Po and ²¹⁰Pb in particles of different sizes in groundwater was determined. Only in one water, with a relatively high NaCl concentration and rich in humus material, was a considerable fraction, about 20%, of both radionuclides found to be present in the soluble form, i.e., passing though the membrane with the smallest pore size. In the other waters only from 1 to 2% of ²¹⁰Po and ²¹⁰Pb was soluble. In most waters the distribution of radionuclides in particles of varying size was quite similar. It is expected that neither ²¹⁰Pb nor especially ²¹⁰Po would form intrinsic precipitates but they would be adsorbed on colloidal minerals and organics. In the ground waters studied practically all uranium (>95%) was in the highly soluble U(VI) form.

When different kinds of treatment methods are used to remove natural radioactivity from drinking water, wastes containing natural radioactivity will be produced. The wastes are in liquid or solid form. Liquid wastes are produced, e.g., when the filters are regenerated or backwashed. Solid wastes are formed, e.g., when various types of filter materials are used. GAC filters emit gamma radiation when they are in use. To gather information on existing national regulations and guides on treatment and disposal of radioactive wastes produced by various water treatment methods, a questionnaire was sent to all the Member Countries of the European Union.

Implications

The data on natural radionuclide levels in ground, drinking and mineral water from 17 European countries and the distribution of uraniferous deposits in Europe enabled the drawing of a European map showing regions which are geologically dominated by basement rocks (especially granite plutons and metamorphic rocks), as the most important areas with potentially elevated levels of natural radionuclides in ground water. It is obvious that, besides these granite-related regions, other smaller areas with high contents of natural radionuclides in ground waters surely exist.

Radon removal systems based on aeration can be designed and installed in many different ways. Average water consumption, maximal momentary consumption and radon concentration in raw water should be considered at least, when the installation is designed. In this study a standard sampling protocol was also developed. The formerly used conventional tests did not provide enough information neither on the effective capacity of the aerators nor on the real removal efficiency.

The study showed that several aeration methods were highly effective in removing radon from water at water works. Removal efficiencies of more than 98% can be achieved, for example, with diffused bubble and packed tower aerators. Most aeration facilities can be constructed to achieve radon removal efficiencies of more than 95% or even more than 99%.

GAC filtration can be considered as an inexpensive and easy way to mitigate moderate concentrations of radon in household water. One of the main concerns, when the GAC filtration is applied in domestic use, is the external gamma radiation that can cause radiation exposure to the residents. An external dose rate can exceed the normal background level even by a factor of one thousand. With a proper shielding, instructions and placement of the unit in a non-living space, elevated doses to the residents, however, can be avoided. In order to minimize exposures, different type of radiation shields were studied. Lead attenuates gamma radiation most efficiently. Bricks and concrete can be applied and a water jacket can be built. However, residential radiation exposures cannot always be eliminated sufficiently, especially when the influent radon activity is high. Therefore, radon removal applying GAC filtration often remains a viable treatment method only when the radon concentration is low.

When considering removal of uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po with activated carbon, the type of carbon should be selected on the basis of its adsorptive properties for these radionuclides. A possible solution for the simultaneous removal of radon and the long-lived radionuclides could be a combined filter based on the carbon-mineral adsorbents. The matrices of these adsorbents could consist of both active carbon and mineral adsorbent.

When using equipment originally designed for iron and manganese removal to remove radon and other radionuclides, the physico-chemical conditions during the filtration should be examined in more detail than in this study. The accurate composition of the masses added into the aeration and filtration equipment needs to be known exactly in order to evaluate the results more reliably. Also the reduction of ²¹⁰Pb and ²¹⁰Po varied largely.

Ion exchange is a proper method for the removal of uranium and ²²⁶Ra. Strong basic anion resins for the removal of uranium and strong acidic cation resins for radium removal performed best. The efficiency for ²¹⁰Pb and ²¹⁰Po varies a lot, since the main proportion of these nuclides is supposed to be particle-bound in natural waters, and therefore no ion exchange process in the real sense, but the adsorption into the resins is responsible for their reduction.

The results showed that when using membrane technology the uranium removal from water at the six tested NF membranes was mainly between 90 and 98%. The high rejection of the uranium compounds is the first sign that uranium can be removed from water by the NF membranes quite effectively. This seems to be valid over a wide range of hydrochemical settings, even in very acidic waters. Beside the six NF membranes, two RO membranes were tested for comparison. As was expected, these membranes rejected uranium (from 98 to 99.5%) more effectively than the NF membranes. In the domestic use the commercial RO-unit removed effectively all radionuclides except radon.

The removal of uranium, ²²⁶Ra, ²¹⁰Pb and ²¹⁰Po from drinking water depends on their speciation. In order to find effective removal methods for these nuclides the knowledge on their speciation in ground water should be known. When selecting methods for removal of ²¹⁰Pb and ²¹⁰Po from ground waters it must be taken into consideration that these radionuclides exist mainly bound in particles in water. In the ground waters from the two drilled wells studied here, practically all uranium was in highly soluble U (VI)-form. Thus, it can be assumed that the oxidation state of uranium has no significant role in removing uranium from drinking water. Instead, the pH of ground water affects on the removal of uranium and should be studied in more detail in the future.

It is recommended that if aeration is used to remove radon from drinking water, the aeration system should be fitted in such a way that the radon released from the water is ventilated into outside air. It is also recommended that the annual dose to inhabitants from external gamma radiation of GAC filter should not exceed 0.1 mSv and that the dose rate at a distance of 1 m from the GAC filter should not exceed 1 µSv/h. To achieve these aims the GAC filter should be equipped with a special shielding to attenuate gamma radiation. The wastes containing natural radioactivity in solid form are also recommended to be discharged into communal dumps and wastes containing natural radioactivity in liquid form to be discharged into sewer.

Coordinator:

Mr Martti Annanmäki (STUK Radiation and Nuclear Safety Authority, FI)

Partners:

Dr Franz Schönhofer (BALUF Federal Institute for Food Control and Research, AT), Prof Dr Hartmut Jungclas (PUMA Philipps University Marburg, Nuclear Chemistry, DE), Mr Reinhard Perfler (IWGA University of Agricultural Sciences Vienna, Institute for Water Provision, Water Ecology and Waste Management, Department for Sanitary Engineering and Water Pollution Control, AT), Mr Lars Mjönes (SSI Swedish Radiation Protection Institute, SE), Prof Dr Rolf-Dieter Wilken (ESWE Institute for Water Research and Water Technology, DE), Prof Dr Timo Jaakkola (HYRL University of Helsinki, Laboratory of Radiochemistry, FI).

Models

Information

Installation